Abstract:While specific training programs are used by trainers to induce particular muscle adaptations in athletes，muscle training is also an important component of progressive rehabilitation to limit the atrophic effect associated withimmobilization，ageing， disease or microgravlty. To optimize these different types of intervention，a good understandingof the response to exercise training and the identification of the mechanisms that determine muscle function are re-quired. While whole muscle performance can be partly determined by adaptations in the firing rate of motorneurons，the pattern of motor unit recruitments or muscle size， the present review will focus speclfically on muscle fiber type ex-pression and the functional adaptations of the contractile apparatus following training. Single skinned fibers have beenshown to be a consistent experimental model to be used for the study of contractile characteristics in vitro，associatedwith determinations of myosin heavy chain isoform expression. The main functional and biochemical properties of hu-man fibers have been studied in detail over the last two decades，but more recently，the skinned fiber model has beenused to investigate the single fiber functional adaptations following different types of solicitations. The aim of this arti-cle is to summarize the responses of the contractile apparatus to specific exercise training programs and to formulatetraining recommendations appiicable to sport sciences and rehabilitation.

Key words: Skinned fibers： Mechanical properties：shortening velocity; Ca2+senaitivity，Electrophoresis

作者簡(jiǎn)介：Smgle flber adaptanons to training

1 introductjon

Human skeletal museles are able t。resond to a largerange 0f functional demands．The functional flebilty whicha110ws the same musele t0 be used for various task5(posturema‘intenance，high precision movements，repeated submaxi‘maIcontractions or powerful maximal contractions)orkinaresftom a sophisticared nervous contr01 and a highly organlzedarrangement of different muscle f|bertypes．These fber typesvary according t0 thelr molecular，metabolic，structural andcontractlle propert‘les and can therefore be grouped actordingto various parameters．The expresslon of different myosinheavy chain(MHC)isoforn2s is a m ajor determl‘nant 0f themechanl’cal characterlstics of mUSCle eells。Hence，a certainfunctional plastlclty is rendered possible via fiber type transl-tl。ons within a muscle，i．e．a(chǎn) modilcaion of the pattern ofMHC IS0forrn express’ion．on the other hand，the eontractl’leproperties of indivIdHal nbers contalnm’g the same MHC iso-forms are also subject to a certai。n degtee of vart’ability andchangesD3．This results most likely ftom differentlal expres—sl’on of other myofibrlllar protei’ns，thkls a110wing a“fine tun-ing”0f one or more charateristics of a given fiber to functi‘on-al demands．

Functlonal plast—lclty 0f different musele f|ber types inhumans has been demonstrated in a humber of studies(34—40)．This feature has been espeelally emphasl‘zed wi‘th proto—c01s uslng changing physi010gical conditi‘ons．Exper’lments arebased on mechanlcaI tests 0f chemlcaUy skinned single muselef-bers that are subsequently analyzed for MHC content．Theresults from the mechanical tests can thus be analyzed sepa—rately for each fiber type．More recently，this model has beenemployed t0 SIudy the charactenstics of the contractile appa—ratus in vari。us population gronps or f。1lowl。ng different‘ypes。f exercise‘interventions．SmaU but systemattc changesin the propertles of the contractlle machinery have been de—tected，which suggests that the si’ngle fIber model iS a power—ful approaeh to reveal subtle changes following exerci’se train～ing.

In depth knowledge about the trainability of different fi-ber types and the impact on whole muscle performance is use-ful to optimize interventions related to sports performanceand rehabilitation. The aim of the present review is thereforeto provide a summary of recent data on single fiber adapta-tions that can be easily interpreted by trainers and ellnleians.The first section provldea an outline of the single fiber modelfor readers not familliarized with this methodology. The nextsection briefly reviews the fiber type transitions detected indifferent training studies. In the third section the main adap-tations of single fibers to specific trainings are presented. Fi-naIly，a series recommendations applicable to sport sciencesand rehabilitation are formulated with special reference to theaging population.

2 Single fiber model

In vitro studies on single muscle fibers have developedrapidly with the possibility to dissect single fiber segmentsfrom needle biopsy samples of human skeletal muscles. Themost commonly studied muscles are lower limb muscles， suchas the soleu ，gastrocnemius，and vastus latera-lis ， while upper limb muscles(deltoidus and biceps brachii) have been less frequently ex-plored. Muscle samples obtained from the b~opsy are immediately bathed in a glycerol solution whereby chemicalskinning of the cells is achieved within days. The collected sample is sectioned longitudinally into small bundles of fibersof approximately 5 mm in length and can be stored for 4 to 5weeks at a temperature of --20\"C.

On the day of the experiment， a single fiber segment ispulled out of the bundle，transferred into relaxing solution de-prived of Caz+ and mounted on an ergometer allowing for fi-ber force and shortening velocity measurements. Fiber con-traction is generated by transferring the segment into a solu-tion with variable Ca+ concentrations. After the completionof the mechanical tests， the fiber segment is removed from theergometer and dissolved in a buffer solution for subsequentMHC isoform determination using gel eleetrophoresis (seeFig. 1). Mechanical properties can thus be analyzed taking in-to account the MHC content of the fiber segment studied. Inaddition，for each study participant the fiber type profile ofthe biopsied muscle is established on the basis of the MHCdetermination of some 100 single muscle fibers.

3 insert Fig 1 about here

Fiber functional properties have been investigated withrespect to several major aspects，based on the methodological approaches used and physiological concerns addressed: (1) determination of fiber cross-sectional area (CSA) and peakforce (Po)， (2) maximal unloaded shortening velocity (Vo)，(3) power output，and (4) determination of Caz+ sensitivity.Figure 1.

MHC isoform determination by 8% acrylamide gel eleetro-phoresis. Lane A depicts the three MHC isoforms I，IIa and llx pres-ent in a \"standard sample\" of human skeletal muscle. Lanes B，C and Drepresent Western blots of the same sample using monodanal anti-MHCI，anti-MHC II and anti MHC I and Ha antibodies，respectively.

Before applying the mechanical tests， the fiber is stretc-hed to a standardized sarcomere length of 2.5 am and evalua-ted with respect to fiber length and diameter， to determine fi-ber CSA and volume. P0 (raN) is recorded as the stable maxi-mal force developed by the fiber while submerged in a solu-tion saturated with Caz+. Specific tension is de-fined as P0/CSA.

V0 is determined via a series of slack tests during whichthe fiber is maximally activated and then quickly shortened toa predefined length. Force initially drops to zero and subse-quently re-increases as the fiber takes up the slack. V0 is cal-culated from the relationship between slack distance and timerequired for force redevelopment. Since the gross values ob-tained depend on the number of sarcomeres in series， V0 isexpressed in fiber lengths per second .

Using the force signal as a feedback control，a series of predefined isotonic force clamps can be performed， yielding each a corresponding value for shortening velocity. The force/ velocity relationship can thus be evaluated， and fiber peak power can be determined. Fiber power is generally expressed in/N FL s and can be normalized with respect to fiber CSA.

The force/Caz+ relationship is determined by measuring submaximal isometric forces developed when the fiber is acti- vated in solutions with different Caz+ concentrations，also ex- pressed as the negative log or pCa. The most commonly used variable to evaluate Caz+ sensitivity is the pCa necessary to a- chieve half-maximal activation of the fiber.

Comparisons between conditions or subject groups clear- ly illustrate the feature of fiber functional plasticity in terms of alterations in the mechanical properties induced by the en- vironment. Figure 2 summarizes the main functional adapta- tions in single type I fibers based on two studies of the group of Widrick and coworkers. The first report is based a cross- sectional comparaison between elite master runners and a group of age-matched controls. The second investigation displays the effects of a 12-week resistance training program in previously untrained young men. Figure 2 not only era-phasizes the considerable variability in terms of single fiberproperties，but also illustrates the concept of training specific-ity detectable at the single fiber level. These issues will befurther addressed in subsequent sections.Figure 2. Relative difference in peak force ( Pe )， cross-sectional area(CSA} ，specific tension ( P0/CSA)， maximal unloaded shortening veloc-ity (V0)，absolnle peak po~' (A. Pwr) and normalized penk power(N. Pwr) In type I fibers from gastrecnemius muscle of master runnersand sedentary people { black }. and from vastus laternlis muscle bnl'oreand after 12 weeks of resistance training ( grey ).

4 Single fiber type transition

A considerable body of work has demonstrated thatMHC isoforms are the main determinants of the functionalproperties of single muscle fibers. MHC isoform expressionhas therefore been frequently used as a molecular marker to i-dentify fiber type. According to the major MHC isoformsfound in adult mammalian skeletal muscle， the following purefiber types can be defined in human skeletal muscle: slow typeI containing MHC I~， fast type IIa containing MHC Ila andfast type IIx containing MHC Ilx (Fig. I).

Muscles mainly implicated in maintaining posture have ahigher proportion of type I fibers. Athletes involved inlong distance events，such as marathon running，have a higher relative content in type I fibers in their leg muscles compared to athletes regularly involved in sprint training . These observations confirm the genera] idea that slow fi- hers are preferentially used for posture control and for slow and repeated contractions， whereas fast fibers provide better performance for rapid and powerful contractions of short du- ration. Additionally， they suggest that muscle cells are dy- namic structures able to change their phenotype under various conditions. The alterations of MHC isoforms have a tendency to follow a general scheme of sequential and reverslble transi- tions from fast-to-slow and slow-to-fast..

5 MHC IoMHC IIaMHC llx (MHC llb)

These transitions between fiber types imply the exista- nee of hybrid fibers in human skeletal muscles， containing both I and IIa or lla and fix MHC isoforma， with rare occur- rences of hybrid I/IIa/IIx fibers. Fiber type transition within a\" muscle is primarily determined by the modifications of the neural impulse patterns delivered to the muscle and by the mechanical loading characteristics. For example， animalstudies have demonstrated that phasic high-frequency stimu-lations，which mimic impulse patterns normally delivered to afast twitch muscle， induce a slow-to-fast transition. On theother hand， chronic low-frequency stimulations mimic impulsepatterns normally delivered to a slow twitch muscle， whichpromotes a fast-to-slow transition.

Exercise training effects on muscle fiber type transitionsare somewhat less radical than those obtained during long-term electrical stimulations in animal models. It is howeverclear that MHC isoform expression can be altered by trainingin humans，despite that there may be a genetic predispositioninducing a gross pattern of muscle MHC profile. It is general-ly accepted that an increase of endurance-type muscle activityinduces a sequential transition from fast to slow muscle fibers(MHC fix MHC IIa MHC I). For example，a short pe-riod of high-intensity endurance training induced a shift fromfast towards slow MHC isoforms in vastus lateralis mus-cler. Another investigation showed that elite marathon run-ners expressed a greater proportion of type I fibers than nor-mally active people (73 versus 51 ). Nevertheless， it ispossible that these individuals had become specialized in longdistance running as a result of an initially greater proportionsof type l fibers.

Some data indicates that not every kind of training in-duces systematically a transition towards more type I fibers. Sprint training has been associated with a higher proportion of MHC IIa and MHC IIx isoforms. A 12-week resistance training program has been shown to induce a decrease in the proportion of muscle fibers containing type IIx MHC isoform (both hybrid type IIa/IIx and type IIx fibers) from 24 to 3 and an increase in the proportion of type IIa fibers from 30 to 55 . A bidirectional transformation (MHC I MHC IIa ~- MHC IIx) has been suggested as a result of combined strength training and interval training. However， resistance training alone does not usually affect the fraction of type I fi- bers. Similarly to the results from resistance training studies，8-weeks plyometric training induced an increase in the proportion of type IIa fibers from 33 to 42 with a tend- ency to decreased proportions of type IIx and hybrid IIa/IIx fibers.

To date， the role of type IIx fibers remains an unresolved question. Type IIx fibers are believed to be the default MHC isoform in humans. These fibers appear in disused mus- cles (following dennervation and or unloading)， are more fati- gable than other fiber types，but seem to have better functlon- al characteristics as they contract faster and generally exhibit greater power output. The transition towards type IIx fibers during chronic muscle unloading could reflect a protection mechanism of the muscle to preserve minimal function. Since plyometric training is characterized by high-velocity and pow-erful contractions， an increase of type IIx fibers could havebeen expected after training. This was， however， not ob-served. D' Antona et al. reported a higher proportion ofMHC fix isoform in vastus lateralis muscles of body builders compared to recreationally active young men (~5). The authors argued that this observation could besport-specific，but an earlier study revealed a lower pro-portion of type llx fibers from biceps brachii of body builderscompared to young sedentary men. It could be speculated thatin humans no feasible movement or exercise training is capa-ble to provoke an increase of the proportion of type llx fibersvia training.

Values represent de mean proportions of muscle fibers contai-ning either type I， IIa or IIx MHC isoforms in different populations orfollowing various training programs. Hybrid fibers contain more thanone MHC isoforms (I/lla or Ila/llx). All results stem from studies inwhich a pool of single fibers were analyzed for MHC isoform contentusing gel electrophoresls. Two of the studies on endurance trainingwere performed on gastrocnemius muscle，while the other datawere

obtained

from

vastus

lateralis

muscle.

Refer-ences are cross-sectional studies and refer-ences are longitudinal studies.

From the previous discussion it appears that differentpopulations and individuals involved in a specific type oftraining are likely to exhibit a fiber type profile that corre-sponds to the functional demands placed upon the musclesconcerned (see Table 1 ). This is particularly marked inahtletes who are highly specialized in a given domain. It is，however，not clear if these subjects had a genetic predisposi-tion that made it possible for them to reach a high level ofperformance， or if systematic long-term training has induced aparticular fiber type profile. Furthermore，it should be kept inmind that changes in whole muscle characteristics observedduring short-term training studies are not always closely re-lated to alterations in relative MHC contentr. Improved con-traction performance can result from uther factors， such asneural activation patterns， muscle metabolic properties or me-chanical characteristics of single fibers. The latter aspect willbe discussed more extensively in the following section.

6 Single fiber ftmetional plasticity

Multiple patterns in functional adaptations of single fi-bers have been illustrated in different populations or in re-sponse to diverse training regimens. These observations haverevealed that fiber functional properties are responsive to themode of muscle activation and thus highly specific to the typeof training.

7 Fiber cross sectional area (CSA) and peak force (Po)

The single fiber model is the only approach to allow thedetermination of maximal fiber force in relation to fiber size.However，it is noteworthy that skinned fibers show some de-gree of swelling after removal of the plasma membrane. A di-rect comparison between the CSA values determined onskinned fibers and those obtained with other techniques istherefore not possible， but can be attempted by using a cor-rection factor (20).

Slow fibers are generally thinner (CSA 5000 gmz)and and develop less force (P0 ~ 0. 8 raN) than fast fibers(CSA ~ 7000/zmz ;P0 ~ 1.2 raN). The difference be-tween fast IIa and IIx fibers is less obvious. Given the differ-ent experimental conditions (temperature， ionic strength， etc~ -') and approaches used in calculating CSA， some disparityappears in fiber P0/CSA vaIues reported by different researchgroups. P0/CSA of slow fibers (lS，3 kN m-z) has beenshown to be lower than P0/CSA of IIa (156 kN ~ m-z) andIIx fibers (170 kN ~ m-z)but this finding was not con-firmed in all studies. Differences in P0/CSA among fastIIa and IIx fibers are even less clear. Thus， P/CSA seemsless affected by the molecular composition (such as MHCisoforms) than other variables (see below).

Some data suggest that fiber P0/CSA may change as aresult of training， but the great majority of studies havefound that alterations in peak force are proportional to chan-ges in fiber size. Therefore， training programsthat induce fiber hypertrophy generally also improve P0. Forexample，resistance training appears to be very efficient to im-prove CSA of type I， IIa and IIa/IIx fibers in young aswell as of type I and type IIa fibers in older people. Fi-ber dimensions were increased in some cases by 3040 fol-lowing a 12-week training program. This fiber hypertro-phy was associated with a proportional increase of Po ，leadingto unaltered P0/CSA. In a recent study， d' Antona et al. re-ported that fast fibers from body builders were － 37--92%larger than those of recreationaIly active individuals，especial-ly type Ilx fibers. However， the great hypertrophy of theirsubjects' vastus lateralis muscles could not be accounted forby mere fiber hypertrophy. The authors argued that the pen-nation angle of muscle fibers could be considered as a possiblesource of discrepancy. They also noted the presence of neo- natal MHC isoforms in the muscles of their body builders， suggesting fiber regeneration and hyperplasia. These ohserva tions must，however，be viewed with caution，since two of the five study participants declared to use anabolic steroids.

Plyometric training， based on explosive， high-load stretch-shortening cycle exercises， has also been proved effec- tive in improving fiber size and force. An 8-week plyometric training programme induced consistent increases in fiber CSA of type I and type II fibers，by 23 and 26 ，respectivelyEz61.In accordance with studies on resistance training， peak force was increased proportionally to fiber hypertrophy. Two in-vestigations have analyzed the effects of or 7 weeks ofsprint training on a cycle ergometer. Contrary to expecta-tions，these studies ailed to observe improvements in P0 ，anda decrease in P0 was even demonstrated in one， These re-suits are surprising since sprint training increased maximumpedal frequency and maximum voluntary contraction of bothknee extensor and plantar flexor musclesEts?. Limited trainingduration may have prevented significant results at the singlefiber level.

Analyzing the effects of endurance training， Widrick etel. (1996a，b) found that CSA of muscle fibers obtained frommaster runners 120 km/week over the preceding 20-25years) where smaller when compared to those from sedentaryaged-matched controls. Nevertheless，Po/CSA was not differ-ent between the two groups. On the other hand，a proportion-al increase of P0 and CSA was found in deltoid type IIa fibersof highly trained collegiate swimmers engaged in a period ofreduced training volume (for 21 days) Harber et areported considerably smaller values for P0 and P0/CSA in cross-country runners after a phase of basicendurance training compared to results from the same lahore-tory gathered from untrained people. The low Po/CSAvalues were， however， comparable after a subsequent phase ofinterval training and a taper period. These results would sug-gest that endurance training tends to decrease the fibers abili-ty to develop force， while a taper period may increase fibersize and force on a short-term basis.

8 Maximal shortening velocity (Vo)

V0 of human fibers at 15\"C varies more than 10-fold be-tween typeI fibers and typeIIx fibers ，with IIa fibers having intermediate values . The existance of hybrid fibers as wellas a certain variability in V0 within a given fiber type makes itpossible to detect a continuum of values over a large functional range. The variability of V0 is relatively greater in type IIfibers than in type I fibers (Fig. 3). Work in which type I IIaand IIx fibers of different human muscles were compared(triceps surae， quadriceps femoris and triceps brachii) sug-gested very similar values for Vo of the same fiber type， regardless of the muscle studied.

Some data from the literature suggests that V0 can begreatly influenced by certain types of training. Widrick etel. ~ found that V0 of type I fibers was 19 higher in mas-ter runners compared to controls and suggested an intrinsicdifference in cross-bridge cycling kinetics in fibers from en-durance-tralned individuals. Nevertheless， V0 of type IIa andIlx fibers was not different between groups. Other longitu-dinal studies have demonstrated that Vo is sensitive to endur-ance training， especially regarding type I fibers.

In contrast， resistance training does not increase V0 ofany fiber type in healthy young men. Gender differenceshave been reported when the effects of resistance trainingwere tested in fibers from older men and women. WhileV0 of type I and type IIa fibers was greatly increased in men，V0 was decreased in type IIa fibers from women after resist-ance training. These data show that in elderly， resistancetraining may well have an influence on fiber V0 and that nenand women may respond differently to this type of training.V~ of type I fibers was reported to he lower in experiencedbody builders when compared to untrained control subjects，while no differences were found for type II fiberscsl. Thus，inagreement with results from studies on resistance training，body building seems inappropriate to increase Vo.

Plyometric training is effective in improving muscle per-formance， especially during explosive movements involvinghigh contraction velocities. In a recent study， Malisoux etel. demonstrated that V was increased in slow and fastfibers of vastus lateralis muscle following 8 weeks of plyo-metric training (Fig. 3). Resoective gains in V. were 18.29and 22 for type I，Ila and Ila/IIxOn the other hand，cycle ergometer sprint training had no effect on V0 ， inspite of increased maximum pedal frequency. The latter effectwas suggested to be caused by improved strength of the ma-jor lower limb muscles.

9 Force/velocity relationship and power

Peak power of type I fibers is ap-proximatively 10-fold lower than that of IIx fibers ，IIa fibers being intermediate . Power-velocity curves illustrate that type I fibersdevelop peak power at a lower optimal contraction velocity than IIa fibers or typeIIx fibers .

Resistance training appears to be an efficient approach toincrease fiber peak power，as the latter was increased in type I(30) and type IIa fibers (42) of healthy young men fol-lowing a 12-week interventlon， In a cross-sectional studyon middle-aged individuals， Shoepe et al reported greaterabsolute peak power for type I， IIa and IIa/IIx fiber from individuals with resistancetraining experience compared to untrained participants. Sinceresistance training does not increase V (see above) the high-er absolute peak power observed in fibers from trained indi-viduals results essentially from fiber hypertrophy and im-provements in Po. The fact that fiber CSA and P， increase inparallel implies that normalized (by CSA) peak power is notinfluenced by this type of exercise. Thus，it appears that re-sistance training induces an increase in the number of crossbridges in parallel， without affecting cross bridge density orintrinsic contractile properties in the fibers of healthy youngindividuals. It is， however， possible that this principle doesnot apply to every population. In elderly men normalized peakpower was improved in type I (57) and type lla fibers(29) as a result of both P0 and V0 increases (36). Resultsobtained from single fibers of elderly women were， however，different， as normalized peak power was not changed aftert raining.

Similarly to resistance training， plyometric training in-duced significant improvements of peak power in slow andfast fibers ，as illustrated in Figure 4. Moreover， the posi-tive effect of this training program on Vo caused normalizedpeak power of type IIa fibers to improve by 9 ， a rather unu-sual finding for healthy young individuals. In parallel， Vow，was enhanced by some 18 n fast fibers. Improvements ofthe contractile properties of single fibers were associated withenhanced performance of the study participants in high powermovements，such as jumping and sprinting.

Absolute peak power was smaller in type I (14) andtype IIa fibers (27 ) from master runners when compared tosedentary controls， essentially as a result of a smaller Po.However，when normalized by CSA，peak power was not dif-ferent between the two groupst. These findings suggestthat fibers from endurance-trained individuals lose their pos-sibility of producing a high power output in favor of increasedcontractile efficiency at elevated velocities. According to theauthors，the higher Vo of the master runners\" type I fibersmay permit these fibers to maintain a higher level of poweroutput during rapid muscle contractions and provide a largercontribution of whole (gastrocnemius) muscle power output.This mechanism would reduce the athlete's dependence onthe more fatigable type IIa fibers and presumably enhance en-durance capacity.

I0 Calcium sensitivity

With the single fiber model，Caz+ sensitivity is evaluated by the force produced by the fiber when exposed to a given submaxlmal Ca+ concentration. Steady state isometric force is recorded in solutions of different Ca~+ concentrations， and the relation between free and force is determined. Care must be taken when interpreting these results. A change in Ca~+ sensitivity alone does not alter the magnitude of con- traction when activation is maximal. The steep part of the force/pCa relationship relates to submaximal activation (Fig 5). In terms of motor unit recruitment，this would correspond to incompletely fused tetanic contractions or submaximal ef- fort. Additionally，due to methodological constraints of the single fiber model，it is difficult to extrapolate functional im- plications from changes in Ca+ sensitivity.

In small mammals， it has been well established that typeI fibers have a higher Ca+ sensitivity than fix fibers，and IIxfibers have higher Caz+ sensitivity than IIa fibers. It is，however，less clear whether or not human fibers differ re-garding Caz sensitivity. For example， Widrick et al. re-ported higher Caz sensitivity in type I fiber at 15\"C (pCa= 6.05 and 6.92 for type I and type IIa， respectively)， andBottinelli et al. reported higher Ca+ sensitivity of type IIaand llx fibers at 12 (pCas=5.66，5.88 and 5.94 for typeI，type IIa and type IIx fibers， respectively). This divergencecould result from the different temperatures at which Ca5+sensitivity was assessed.

Caz+ sensitivity has been less frequently investigatedthan other fiber characteristics. This could be explained inpart by the fact that a number of studies failed to show aneffect of training in the force/pCa relationship. For example，no change was observed after sprint training for type I， IIaand IIa/llx fiber，and no difference was observed betweenmaster runners and sedentary people for type I and type finfibers . The effects of resistance training on force/pCa rela-tionships have never been investigated， except in olderwomen. In this population，Caz+ sensitivity was indeed in-creased in type I fibers. This finding was in agreement withthe increase in Caz+ sensitivity of type I fibers followingplyometrie training. Neither resistance trainingr norplyometrie trsining had any effects on the Caz+ sensitivityof type Ila and IIa/Ilx fibers. These observations imply thattype I fibers are more responsive to physical activity andtraining regarding changes in Caz+ sensitivity.

11 Recommendations and conclusion

Professionals in the field of exercise training and rehabil-itation have an interest in understanding the precise effectsinduced by different training paradigms. Investigations thathave evaluated performance characteristics in association withsingle muscle fiber analyses have demonstrated that wholemuscle performance depends to a considerable extent on theexpression pattern and the mechanical properties of the dif-ferent fiber types. Intervention studies have shown that train-ing effects can be observed after 8 to 1 weeks of trainingmade up of 2 to 4 weekly sessions. However， from the pre-ceeding sections it appears that structural and functional ad-aptations observed at the single fiber level are highly specificto the type of training administered. Furthermore，the trainingeffects have been found to be population-specific， in that eld-erly individuals may react differently to a given training stim-ulus.

Resistance training involves short sets of low-velocitycontractions against high loads， generally －800 of one-repe-tition maximal (1-RM) contraction. Most studies presented a-bove have applied this definition in their experimental design.However， the modalities of resistance training are often ad-justed with respect to the aim followed (athletic training，re-habilitation，etc. ) and the population concerned， which callsfor some care in the interpretation of the presented results.The major response of relatively young sedentary subjects toresistance training is a proportional increase of Po and fiberCSA， inducing an improvement of absolute peak power. Giventhat V0 does not change， the increase of power is the result offiber hypertrophy and enhanced force. Thus， when normalizedby CSA， peak power is not influenced by resistance trainingand Vopt remains unchanged. These results imply that resist-ance training is efficient to improve muscle functional per-formance for tasks requiring high forces and low velocities.However，powerful movements performed at relatively highvelocities are not likely to benefit from such training.

Body building，also called hypertrophic heavy resistancetraining，ls characterized by very short sets of powerful con-tractions with nearly maximal loads (90-95of 1-RM).This type of exercise seems to favor the expression of themost powerful fiber type (IIx)，and to induce fiber hypertro-phy and hyperplasia. However，given that maximal shorteningvelocity of single fibers is unchanged or even decreased， thistraining paradigm is not appropriate for activities involvingfast or explosive muscle contractions.

Plyometric training is characterized by high-intensity，high-velocity contractions， mostly performed without over-weight. It is a particularly interesting method to increase per-formance during powerful movements (e. g. jumping or sprin-ting). Functional improvements can partly be attributed toenhanced force， velocity and， consequently， power of slow andfast muscle fibers. An important feature to obtain the desired improvements with this training method is that movementsare executed as fast as possible and at maximal intensity.

Sprint training is characterized by exercises involvingshort duration， high power and high velocity contractions.Only a few data are available in the literature， and the studies in this domain were unable to link improvements of wholemuscle performance to changes in single fiber properties. Itshould be noted， however， that the training periods used inthose studies were rather short. Better sprint performancecould be attributed to improved coordination or optimizationof neural control， leading to enhanced force production. Todate，no information is available on the effects of sprint train-ing on single fiber power output. Furhtermore， future investi-gations should focus on the effects of sprint training in run-ning.

Endurance training is characterized by high-volume， lowforce contractions. The higher V0 observed in endurancetrained individuals can be associated with a lower fiber poweroutput. While this adaptation is compatible with an increaseof endurance capacity，the muscle's ability to develop powerin tasks involving high loads (and， therefore， low contractionvelocities) is consequently reduced. The effects of a taper pe-riod preceding competition have been recently investigated，but the results are contradictory. In one study on endurancerunners V0 and peak power of type I fibers were decreasedafter a phase of interval training followed by a taper period.These results are in opposition to the increases in fiber force，velocity and peak power found in swimmers after a 3-weektaper phase. Possible explanations for these discrepanciescould stem from differences in the muscles investigated (gas-trucnemius in the former and deltoidus in the latter study)，the population studied or the training program applied. Clear-ly， the effects of interval training and taper period are insuffi-ciently understood to date and deserve further attention， tak-ing into account the specificity of the sports discipline.

Exercise training in elderly deserves some special con-siderations. Ageing has been associated with a decrease in theproportion of type lI fibers， a decrease in fiber CSA and an al-teration of contractile properties， such as P0， P0/CSA andVo. Type If fibers appear to be more affected， as agreater decrease in CSA was observed in these fibers，leadingto an increase in the relative muscle area occupied by type Ifibers. Sprint training has been shown to reduce age effects onsome of the single fiber mechanical characteristics. Korhonenet al. ~2o~ reported that in 70-84 year old male sprinter runnersCSA of type I fibers was maintained. Additionally， P0/CSAand V0 of type I and type IIa fibers were unaffected in themaster-aged group when compared to young athletes (18－33years). Therefore， sprint training appears to be an efficientstimulus to maintain muscle fiber structure and force produc-tion. Trappe and coworkers investigated the effects of resist-anee training on the functional properties of single muscle fi-bers in older people of 75 years. In men and women，reaistanee training increased muscle cell size，strength and ab-solute peak power in type I and type IIa fibers. Additionally，V0 and normalized peak power was increased in type I andtype fin fibers in men，but not in women. These results sug-gent that men and women respond dlfferently to resistancetraining. Moreover， it could be put forward that single fibercontractile properties are sensitive to a change of physical ac-tivity rather than ageing. In summary， it is important to en-courage older people to be active. While explosive-type con-tractions and overload exercises seem most efficient to reducethe age-related loss of muscle function， a conservative ap-proach should be favored when applying these exercise mo-dalities with elderly individuals.

In conclusion， a number of studies have demonstratedthat enhanced whole muscle performances induced by trainingare associated with parallel functional improvements in singlemuscle fibers. The concept of training specificity is thus verywell reflected at the cellular level. Although the single fibermodel does not provide a straightforward picture of the fibercharacteristics in physiological conditions， this approachmakes it possible to detect subtle changes in the contractileapparatus that help understand muscle performance.